Cortisol, inflammation and allostatic load

The response that works

A threat appears. Within a second, sympathetic outflow rises and vagal restraint on the heart withdraws: heart rate climbs, blood pressure rises, blood is redirected to muscle, attention narrows. Within minutes, the hypothalamic-pituitary-adrenal axis releases cortisol, which mobilises glucose and holds the response steady while it is needed.

Then — and this is the part that matters — cortisol feeds back on the hypothalamus and pituitary and shuts the axis down. The system is designed to terminate itself. Everything about the acute response is adaptive, including the fact that it ends.

Chronic stress is not a stronger version of this. It is a version in which the termination step stops working properly.

When the rhythm flattens

Cortisol is not just a stress hormone; it is a rhythm. In a healthy adult it rises sharply in the 30 minutes after waking, peaks, and then falls steadily across the day to a trough around midnight. The steepness of that decline — the diurnal slope — is a better marker of regulation than any single measurement.

Under sustained stress, the slope flattens. Morning peaks blunt, evening levels stay elevated, and the difference between night and day narrows. Meta-analytic work links flatter diurnal slopes to worse outcomes across a broad range of physical and mental health measures, including immune function, cardiovascular disease and mortality.

A flattened slope also explains one of the most common experiences people report on the way into burnout: tired in the morning, wired at night. That is not a metaphor. It is the shape of the curve.

The inflammatory shift

Cortisol normally restrains inflammation. When the axis is chronically activated, immune cells become less responsive to that restraint — a state usually described as glucocorticoid resistance — and pro-inflammatory signalling drifts upward. The markers that move are the familiar ones: interleukin-6, tumour necrosis factor alpha, C-reactive protein.

None of this reaches the level of an acute illness. It is low-grade, systemic and persistent, and it is now considered a common mechanism underlying a striking share of age-related disease — cardiovascular disease, type 2 diabetes, neurodegeneration, and depression among them.

Allostatic load

Bruce McEwen's 1998 framing remains the most useful summary. Homeostasis is holding a variable steady; allostasis is achieving stability through change — raising blood pressure, mobilising glucose, suppressing digestion — in order to meet a demand. Allostasis is what keeps you alive under load.

Allostatic load is the price of running it too long. McEwen described four patterns that generate it: too-frequent activation, failure to shut off after the stressor ends, failure to habituate to a repeated stressor, and an inadequate response in one system that forces another to compensate.

Notice that three of those four are about regulation rather than exposure. The problem is rarely that something demanding happened. It is that the response to it did not resolve.

Why this looks like accelerated aging

Put the pieces together and the profile of long-term stress overlaps substantially with the profile of aging: flatter hormonal rhythms, higher inflammatory tone, reduced heart-rate variability, poorer sleep quality, impaired glucose handling, and reduced immune competence.

Work on cellular markers points the same way. In a widely cited 2004 study, women reporting the highest levels of perceived stress had shorter telomeres and lower telomerase activity than low-stress controls — a difference the authors estimated at around a decade of additional cellular aging. Telomere research has become more cautious since, and effect sizes vary between studies, but the direction of the finding has held up across a large literature.

This is the reasoning behind the resilience-age readout in the Calmspan test. It is a translation, not a measurement — but the thing it is translating is real.

Duration beats intensity

The most practically important finding in this whole area is that peak reactivity predicts far less than recovery time. Two people can show identical cardiovascular spikes during a laboratory stressor and diverge sharply in outcomes, because one returns to baseline in ten minutes and the other is still elevated an hour later.

What extends the response, more than anything else, is cognitive: rumination, anticipation, and mental rehearsal keep the physiological signal running in the absence of the stressor. The argument lasted four minutes. The tenth replay of it is doing the damage.

What this means in practice

• Optimise for recovery, not for calm. A demanding life with clean recovery is physiologically cheaper than a quiet life spent ruminating.
• Protect the rhythm. Fixed wake time, morning light, and a genuine wind-down do more for the cortisol slope than any intervention aimed at the peak.
• Treat rumination as the target. It is the mechanism that converts a short stressor into a long one, and it responds to specific techniques — scheduled worry time, affect labelling, cognitive behavioural approaches.
• Interrupt physiologically. Slow breathing shortens the tail of the response directly, in minutes; see the resonance breathing guide.